| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 15.0, | |
| "eval_steps": 500, | |
| "global_step": 23445, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.06397952655150352, | |
| "grad_norm": 1.3757662773132324, | |
| "learning_rate": 0.0002987204094689699, | |
| "loss": 2.6581, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.12795905310300704, | |
| "grad_norm": 1.3149511814117432, | |
| "learning_rate": 0.0002974408189379398, | |
| "loss": 2.2938, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.19193857965451055, | |
| "grad_norm": 1.3127187490463257, | |
| "learning_rate": 0.00029617402431222006, | |
| "loss": 2.2174, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.2559181062060141, | |
| "grad_norm": 1.1597858667373657, | |
| "learning_rate": 0.00029489443378119, | |
| "loss": 2.1893, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.3198976327575176, | |
| "grad_norm": 1.3044805526733398, | |
| "learning_rate": 0.00029361484325015995, | |
| "loss": 2.173, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.3838771593090211, | |
| "grad_norm": 1.2641485929489136, | |
| "learning_rate": 0.00029233525271912987, | |
| "loss": 2.1609, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.44785668586052463, | |
| "grad_norm": 1.2622572183609009, | |
| "learning_rate": 0.0002910556621880998, | |
| "loss": 2.1425, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.5118362124120281, | |
| "grad_norm": 1.181531548500061, | |
| "learning_rate": 0.0002897760716570697, | |
| "loss": 2.1016, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.5758157389635317, | |
| "grad_norm": 1.1225402355194092, | |
| "learning_rate": 0.0002884964811260397, | |
| "loss": 2.087, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.6397952655150352, | |
| "grad_norm": 1.2168927192687988, | |
| "learning_rate": 0.00028721689059500955, | |
| "loss": 2.1028, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.7037747920665387, | |
| "grad_norm": 1.1682486534118652, | |
| "learning_rate": 0.0002859373000639795, | |
| "loss": 2.0909, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.7677543186180422, | |
| "grad_norm": 1.1344854831695557, | |
| "learning_rate": 0.00028465770953294944, | |
| "loss": 2.0896, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.8317338451695457, | |
| "grad_norm": 1.1923776865005493, | |
| "learning_rate": 0.00028337811900191936, | |
| "loss": 2.0488, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.8957133717210493, | |
| "grad_norm": 1.3097106218338013, | |
| "learning_rate": 0.0002820985284708893, | |
| "loss": 2.0488, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.9596928982725528, | |
| "grad_norm": 1.1138262748718262, | |
| "learning_rate": 0.0002808189379398592, | |
| "loss": 2.0354, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.0236724248240563, | |
| "grad_norm": 1.2939698696136475, | |
| "learning_rate": 0.00027953934740882917, | |
| "loss": 1.9789, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.0876519513755598, | |
| "grad_norm": 1.2986066341400146, | |
| "learning_rate": 0.0002782597568777991, | |
| "loss": 1.9117, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.1516314779270633, | |
| "grad_norm": 1.36199152469635, | |
| "learning_rate": 0.000276980166346769, | |
| "loss": 1.9463, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.2156110044785668, | |
| "grad_norm": 1.1844524145126343, | |
| "learning_rate": 0.00027570057581573893, | |
| "loss": 1.9163, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 1.2795905310300704, | |
| "grad_norm": 1.3748372793197632, | |
| "learning_rate": 0.0002744209852847089, | |
| "loss": 1.9157, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.3435700575815739, | |
| "grad_norm": 1.3857694864273071, | |
| "learning_rate": 0.00027314139475367877, | |
| "loss": 1.9234, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 1.4075495841330774, | |
| "grad_norm": 1.244759202003479, | |
| "learning_rate": 0.00027186180422264874, | |
| "loss": 1.9202, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 1.471529110684581, | |
| "grad_norm": 1.3650200366973877, | |
| "learning_rate": 0.00027058221369161866, | |
| "loss": 1.9262, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 1.5355086372360844, | |
| "grad_norm": 1.3304725885391235, | |
| "learning_rate": 0.0002693026231605886, | |
| "loss": 1.9225, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 1.599488163787588, | |
| "grad_norm": 1.2908028364181519, | |
| "learning_rate": 0.0002680230326295585, | |
| "loss": 1.9448, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.6634676903390915, | |
| "grad_norm": 1.4424597024917603, | |
| "learning_rate": 0.00026674344209852847, | |
| "loss": 1.9327, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 1.727447216890595, | |
| "grad_norm": 1.3038992881774902, | |
| "learning_rate": 0.0002654638515674984, | |
| "loss": 1.9156, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 1.7914267434420985, | |
| "grad_norm": 1.2507870197296143, | |
| "learning_rate": 0.0002641842610364683, | |
| "loss": 1.9111, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 1.855406269993602, | |
| "grad_norm": 1.3444691896438599, | |
| "learning_rate": 0.00026290467050543823, | |
| "loss": 1.9101, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 1.9193857965451055, | |
| "grad_norm": 1.3593779802322388, | |
| "learning_rate": 0.00026162507997440815, | |
| "loss": 1.9322, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.983365323096609, | |
| "grad_norm": 1.3340197801589966, | |
| "learning_rate": 0.0002603454894433781, | |
| "loss": 1.9361, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 2.0473448496481126, | |
| "grad_norm": 1.5682820081710815, | |
| "learning_rate": 0.000259065898912348, | |
| "loss": 1.8007, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 2.111324376199616, | |
| "grad_norm": 1.5191752910614014, | |
| "learning_rate": 0.00025778630838131796, | |
| "loss": 1.7626, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 2.1753039027511196, | |
| "grad_norm": 1.614561915397644, | |
| "learning_rate": 0.0002565067178502879, | |
| "loss": 1.7621, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 2.239283429302623, | |
| "grad_norm": 1.4934717416763306, | |
| "learning_rate": 0.00025522712731925785, | |
| "loss": 1.7753, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 2.3032629558541267, | |
| "grad_norm": 1.6563340425491333, | |
| "learning_rate": 0.0002539475367882277, | |
| "loss": 1.771, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 2.36724248240563, | |
| "grad_norm": 1.502406120300293, | |
| "learning_rate": 0.0002526679462571977, | |
| "loss": 1.7823, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 2.4312220089571337, | |
| "grad_norm": 1.6300321817398071, | |
| "learning_rate": 0.0002513883557261676, | |
| "loss": 1.7792, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 2.495201535508637, | |
| "grad_norm": 1.5266133546829224, | |
| "learning_rate": 0.00025010876519513753, | |
| "loss": 1.7994, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 2.5591810620601407, | |
| "grad_norm": 1.5854878425598145, | |
| "learning_rate": 0.00024882917466410745, | |
| "loss": 1.7947, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.6231605886116443, | |
| "grad_norm": 1.4971704483032227, | |
| "learning_rate": 0.00024754958413307737, | |
| "loss": 1.8, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 2.6871401151631478, | |
| "grad_norm": 1.4385017156600952, | |
| "learning_rate": 0.00024626999360204734, | |
| "loss": 1.8041, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 2.7511196417146513, | |
| "grad_norm": 1.526516079902649, | |
| "learning_rate": 0.00024499040307101726, | |
| "loss": 1.8172, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 2.815099168266155, | |
| "grad_norm": 1.5419269800186157, | |
| "learning_rate": 0.00024371081253998718, | |
| "loss": 1.7861, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 2.8790786948176583, | |
| "grad_norm": 1.6025625467300415, | |
| "learning_rate": 0.00024243122200895713, | |
| "loss": 1.8099, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.943058221369162, | |
| "grad_norm": 1.6059303283691406, | |
| "learning_rate": 0.00024115163147792705, | |
| "loss": 1.8057, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 3.0070377479206654, | |
| "grad_norm": 1.4824891090393066, | |
| "learning_rate": 0.00023987204094689697, | |
| "loss": 1.7666, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 3.071017274472169, | |
| "grad_norm": 1.7200456857681274, | |
| "learning_rate": 0.00023859245041586688, | |
| "loss": 1.5827, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 3.1349968010236724, | |
| "grad_norm": 1.7883553504943848, | |
| "learning_rate": 0.00023731285988483683, | |
| "loss": 1.6154, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 3.198976327575176, | |
| "grad_norm": 1.7612642049789429, | |
| "learning_rate": 0.00023604606525911705, | |
| "loss": 1.6331, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 3.2629558541266794, | |
| "grad_norm": 1.8273234367370605, | |
| "learning_rate": 0.000234766474728087, | |
| "loss": 1.6489, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 3.326935380678183, | |
| "grad_norm": 1.9144798517227173, | |
| "learning_rate": 0.00023348688419705691, | |
| "loss": 1.6451, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 3.3909149072296865, | |
| "grad_norm": 1.7729228734970093, | |
| "learning_rate": 0.00023220729366602686, | |
| "loss": 1.6504, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 3.45489443378119, | |
| "grad_norm": 1.7989096641540527, | |
| "learning_rate": 0.00023092770313499675, | |
| "loss": 1.6774, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 3.5188739603326935, | |
| "grad_norm": 1.9125975370407104, | |
| "learning_rate": 0.0002296481126039667, | |
| "loss": 1.6734, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 3.582853486884197, | |
| "grad_norm": 1.764624834060669, | |
| "learning_rate": 0.00022836852207293665, | |
| "loss": 1.6829, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 3.6468330134357005, | |
| "grad_norm": 1.8827048540115356, | |
| "learning_rate": 0.0002270889315419066, | |
| "loss": 1.6917, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 3.710812539987204, | |
| "grad_norm": 1.7637380361557007, | |
| "learning_rate": 0.00022580934101087648, | |
| "loss": 1.6957, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 3.7747920665387076, | |
| "grad_norm": 1.9357694387435913, | |
| "learning_rate": 0.00022452975047984643, | |
| "loss": 1.7071, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 3.838771593090211, | |
| "grad_norm": 1.8760075569152832, | |
| "learning_rate": 0.00022325015994881638, | |
| "loss": 1.689, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 3.9027511196417146, | |
| "grad_norm": 1.83319890499115, | |
| "learning_rate": 0.0002219705694177863, | |
| "loss": 1.7135, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 3.966730646193218, | |
| "grad_norm": 1.8324424028396606, | |
| "learning_rate": 0.00022069097888675622, | |
| "loss": 1.7052, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 4.030710172744722, | |
| "grad_norm": 1.772605538368225, | |
| "learning_rate": 0.00021941138835572613, | |
| "loss": 1.6098, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 4.094689699296225, | |
| "grad_norm": 1.936546802520752, | |
| "learning_rate": 0.00021813179782469608, | |
| "loss": 1.4814, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 4.158669225847729, | |
| "grad_norm": 2.055730104446411, | |
| "learning_rate": 0.000216852207293666, | |
| "loss": 1.5106, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 4.222648752399232, | |
| "grad_norm": 2.3238749504089355, | |
| "learning_rate": 0.00021557261676263592, | |
| "loss": 1.5136, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 4.286628278950736, | |
| "grad_norm": 1.9317381381988525, | |
| "learning_rate": 0.00021429302623160587, | |
| "loss": 1.5307, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 4.350607805502239, | |
| "grad_norm": 2.057237386703491, | |
| "learning_rate": 0.0002130134357005758, | |
| "loss": 1.5524, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 4.414587332053743, | |
| "grad_norm": 2.2331132888793945, | |
| "learning_rate": 0.0002117338451695457, | |
| "loss": 1.5674, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 4.478566858605246, | |
| "grad_norm": 2.0981638431549072, | |
| "learning_rate": 0.00021046705054382595, | |
| "loss": 1.5627, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 4.54254638515675, | |
| "grad_norm": 2.1681783199310303, | |
| "learning_rate": 0.0002091874600127959, | |
| "loss": 1.5615, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 4.606525911708253, | |
| "grad_norm": 2.0720126628875732, | |
| "learning_rate": 0.00020790786948176584, | |
| "loss": 1.5598, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 4.670505438259757, | |
| "grad_norm": 1.9248210191726685, | |
| "learning_rate": 0.00020662827895073573, | |
| "loss": 1.5712, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 4.73448496481126, | |
| "grad_norm": 1.9172708988189697, | |
| "learning_rate": 0.00020534868841970568, | |
| "loss": 1.5629, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 4.798464491362764, | |
| "grad_norm": 2.141303539276123, | |
| "learning_rate": 0.00020406909788867563, | |
| "loss": 1.5871, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 4.862444017914267, | |
| "grad_norm": 1.9447873830795288, | |
| "learning_rate": 0.00020278950735764552, | |
| "loss": 1.6006, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 4.926423544465771, | |
| "grad_norm": 2.089735984802246, | |
| "learning_rate": 0.00020150991682661547, | |
| "loss": 1.5995, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 4.990403071017274, | |
| "grad_norm": 2.056344985961914, | |
| "learning_rate": 0.00020023032629558539, | |
| "loss": 1.6122, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 5.054382597568778, | |
| "grad_norm": 2.0974326133728027, | |
| "learning_rate": 0.00019895073576455533, | |
| "loss": 1.4254, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 5.1183621241202815, | |
| "grad_norm": 2.250195264816284, | |
| "learning_rate": 0.00019767114523352525, | |
| "loss": 1.3966, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 5.182341650671785, | |
| "grad_norm": 2.278472900390625, | |
| "learning_rate": 0.00019639155470249517, | |
| "loss": 1.4051, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 5.2463211772232885, | |
| "grad_norm": 2.3049135208129883, | |
| "learning_rate": 0.00019511196417146512, | |
| "loss": 1.4147, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 5.310300703774792, | |
| "grad_norm": 2.423823833465576, | |
| "learning_rate": 0.00019383237364043506, | |
| "loss": 1.4283, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 5.3742802303262955, | |
| "grad_norm": 2.4221420288085938, | |
| "learning_rate": 0.00019255278310940495, | |
| "loss": 1.4538, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 5.438259756877799, | |
| "grad_norm": 2.3886525630950928, | |
| "learning_rate": 0.0001912731925783749, | |
| "loss": 1.4435, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 5.502239283429303, | |
| "grad_norm": 2.4025745391845703, | |
| "learning_rate": 0.00018999360204734485, | |
| "loss": 1.4512, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 5.566218809980806, | |
| "grad_norm": 2.312255382537842, | |
| "learning_rate": 0.00018871401151631477, | |
| "loss": 1.4644, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 5.63019833653231, | |
| "grad_norm": 2.2402162551879883, | |
| "learning_rate": 0.00018743442098528469, | |
| "loss": 1.4844, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 5.694177863083813, | |
| "grad_norm": 2.3729898929595947, | |
| "learning_rate": 0.00018615483045425463, | |
| "loss": 1.4749, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 5.758157389635317, | |
| "grad_norm": 2.137364149093628, | |
| "learning_rate": 0.00018488803582853488, | |
| "loss": 1.4875, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 5.82213691618682, | |
| "grad_norm": 2.404008626937866, | |
| "learning_rate": 0.00018360844529750477, | |
| "loss": 1.4954, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 5.886116442738324, | |
| "grad_norm": 2.167051076889038, | |
| "learning_rate": 0.00018232885476647472, | |
| "loss": 1.4923, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 5.950095969289827, | |
| "grad_norm": 2.2124693393707275, | |
| "learning_rate": 0.00018104926423544464, | |
| "loss": 1.4972, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 6.014075495841331, | |
| "grad_norm": 2.4574244022369385, | |
| "learning_rate": 0.00017976967370441458, | |
| "loss": 1.4535, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 6.078055022392834, | |
| "grad_norm": 2.2989892959594727, | |
| "learning_rate": 0.0001784900831733845, | |
| "loss": 1.272, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 6.142034548944338, | |
| "grad_norm": 2.8099708557128906, | |
| "learning_rate": 0.00017721049264235442, | |
| "loss": 1.3033, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 6.206014075495841, | |
| "grad_norm": 2.516444206237793, | |
| "learning_rate": 0.00017593090211132437, | |
| "loss": 1.3066, | |
| "step": 9700 | |
| }, | |
| { | |
| "epoch": 6.269993602047345, | |
| "grad_norm": 2.617293357849121, | |
| "learning_rate": 0.0001746513115802943, | |
| "loss": 1.3263, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 6.333973128598848, | |
| "grad_norm": 2.6873817443847656, | |
| "learning_rate": 0.0001733717210492642, | |
| "loss": 1.3415, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 6.397952655150352, | |
| "grad_norm": 2.558847427368164, | |
| "learning_rate": 0.00017209213051823415, | |
| "loss": 1.3599, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 6.461932181701855, | |
| "grad_norm": 2.6037933826446533, | |
| "learning_rate": 0.0001708125399872041, | |
| "loss": 1.3486, | |
| "step": 10100 | |
| }, | |
| { | |
| "epoch": 6.525911708253359, | |
| "grad_norm": 2.470381259918213, | |
| "learning_rate": 0.000169532949456174, | |
| "loss": 1.3639, | |
| "step": 10200 | |
| }, | |
| { | |
| "epoch": 6.589891234804862, | |
| "grad_norm": 2.6497058868408203, | |
| "learning_rate": 0.00016825335892514394, | |
| "loss": 1.365, | |
| "step": 10300 | |
| }, | |
| { | |
| "epoch": 6.653870761356366, | |
| "grad_norm": 2.8465592861175537, | |
| "learning_rate": 0.00016697376839411388, | |
| "loss": 1.3754, | |
| "step": 10400 | |
| }, | |
| { | |
| "epoch": 6.717850287907869, | |
| "grad_norm": 2.4625210762023926, | |
| "learning_rate": 0.0001656941778630838, | |
| "loss": 1.3743, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 6.781829814459373, | |
| "grad_norm": 2.633486747741699, | |
| "learning_rate": 0.00016441458733205372, | |
| "loss": 1.3786, | |
| "step": 10600 | |
| }, | |
| { | |
| "epoch": 6.8458093410108765, | |
| "grad_norm": 2.799623727798462, | |
| "learning_rate": 0.00016313499680102367, | |
| "loss": 1.3837, | |
| "step": 10700 | |
| }, | |
| { | |
| "epoch": 6.90978886756238, | |
| "grad_norm": 2.4427671432495117, | |
| "learning_rate": 0.0001618554062699936, | |
| "loss": 1.4229, | |
| "step": 10800 | |
| }, | |
| { | |
| "epoch": 6.9737683941138835, | |
| "grad_norm": 2.680490016937256, | |
| "learning_rate": 0.00016057581573896353, | |
| "loss": 1.3956, | |
| "step": 10900 | |
| }, | |
| { | |
| "epoch": 7.037747920665387, | |
| "grad_norm": 2.6439030170440674, | |
| "learning_rate": 0.00015929622520793343, | |
| "loss": 1.2917, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 7.1017274472168905, | |
| "grad_norm": 2.618804454803467, | |
| "learning_rate": 0.00015802943058221367, | |
| "loss": 1.2111, | |
| "step": 11100 | |
| }, | |
| { | |
| "epoch": 7.165706973768394, | |
| "grad_norm": 2.7387425899505615, | |
| "learning_rate": 0.00015674984005118362, | |
| "loss": 1.2036, | |
| "step": 11200 | |
| }, | |
| { | |
| "epoch": 7.229686500319898, | |
| "grad_norm": 2.7440712451934814, | |
| "learning_rate": 0.00015547024952015354, | |
| "loss": 1.2316, | |
| "step": 11300 | |
| }, | |
| { | |
| "epoch": 7.293666026871401, | |
| "grad_norm": 2.687152147293091, | |
| "learning_rate": 0.00015419065898912346, | |
| "loss": 1.2214, | |
| "step": 11400 | |
| }, | |
| { | |
| "epoch": 7.357645553422905, | |
| "grad_norm": 2.7659342288970947, | |
| "learning_rate": 0.0001529110684580934, | |
| "loss": 1.2316, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 7.421625079974408, | |
| "grad_norm": 2.877668857574463, | |
| "learning_rate": 0.00015163147792706335, | |
| "loss": 1.2567, | |
| "step": 11600 | |
| }, | |
| { | |
| "epoch": 7.485604606525912, | |
| "grad_norm": 2.554941415786743, | |
| "learning_rate": 0.00015035188739603324, | |
| "loss": 1.2661, | |
| "step": 11700 | |
| }, | |
| { | |
| "epoch": 7.549584133077415, | |
| "grad_norm": 2.9956531524658203, | |
| "learning_rate": 0.0001490722968650032, | |
| "loss": 1.2855, | |
| "step": 11800 | |
| }, | |
| { | |
| "epoch": 7.613563659628919, | |
| "grad_norm": 2.6620912551879883, | |
| "learning_rate": 0.00014779270633397313, | |
| "loss": 1.288, | |
| "step": 11900 | |
| }, | |
| { | |
| "epoch": 7.677543186180422, | |
| "grad_norm": 2.637007713317871, | |
| "learning_rate": 0.00014651311580294305, | |
| "loss": 1.2665, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 7.741522712731926, | |
| "grad_norm": 2.8605268001556396, | |
| "learning_rate": 0.00014523352527191297, | |
| "loss": 1.2897, | |
| "step": 12100 | |
| }, | |
| { | |
| "epoch": 7.805502239283429, | |
| "grad_norm": 2.8997604846954346, | |
| "learning_rate": 0.0001439539347408829, | |
| "loss": 1.3052, | |
| "step": 12200 | |
| }, | |
| { | |
| "epoch": 7.869481765834933, | |
| "grad_norm": 2.889934778213501, | |
| "learning_rate": 0.00014267434420985284, | |
| "loss": 1.2934, | |
| "step": 12300 | |
| }, | |
| { | |
| "epoch": 7.933461292386436, | |
| "grad_norm": 2.7243242263793945, | |
| "learning_rate": 0.00014139475367882276, | |
| "loss": 1.3035, | |
| "step": 12400 | |
| }, | |
| { | |
| "epoch": 7.99744081893794, | |
| "grad_norm": 2.7266464233398438, | |
| "learning_rate": 0.00014011516314779268, | |
| "loss": 1.3032, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 8.061420345489443, | |
| "grad_norm": 2.5178606510162354, | |
| "learning_rate": 0.00013883557261676262, | |
| "loss": 1.125, | |
| "step": 12600 | |
| }, | |
| { | |
| "epoch": 8.125399872040948, | |
| "grad_norm": 2.5766046047210693, | |
| "learning_rate": 0.00013755598208573254, | |
| "loss": 1.1352, | |
| "step": 12700 | |
| }, | |
| { | |
| "epoch": 8.18937939859245, | |
| "grad_norm": 2.6406075954437256, | |
| "learning_rate": 0.0001362763915547025, | |
| "loss": 1.1377, | |
| "step": 12800 | |
| }, | |
| { | |
| "epoch": 8.253358925143955, | |
| "grad_norm": 2.6050360202789307, | |
| "learning_rate": 0.0001349968010236724, | |
| "loss": 1.1539, | |
| "step": 12900 | |
| }, | |
| { | |
| "epoch": 8.317338451695457, | |
| "grad_norm": 3.0384466648101807, | |
| "learning_rate": 0.00013371721049264235, | |
| "loss": 1.1446, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 8.381317978246962, | |
| "grad_norm": 2.728938102722168, | |
| "learning_rate": 0.00013245041586692257, | |
| "loss": 1.154, | |
| "step": 13100 | |
| }, | |
| { | |
| "epoch": 8.445297504798464, | |
| "grad_norm": 2.723478078842163, | |
| "learning_rate": 0.00013117082533589252, | |
| "loss": 1.1583, | |
| "step": 13200 | |
| }, | |
| { | |
| "epoch": 8.509277031349969, | |
| "grad_norm": 2.9142537117004395, | |
| "learning_rate": 0.00012989123480486244, | |
| "loss": 1.1697, | |
| "step": 13300 | |
| }, | |
| { | |
| "epoch": 8.573256557901471, | |
| "grad_norm": 3.0678508281707764, | |
| "learning_rate": 0.00012861164427383238, | |
| "loss": 1.1759, | |
| "step": 13400 | |
| }, | |
| { | |
| "epoch": 8.637236084452976, | |
| "grad_norm": 2.6915736198425293, | |
| "learning_rate": 0.0001273320537428023, | |
| "loss": 1.18, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 8.701215611004478, | |
| "grad_norm": 2.8362019062042236, | |
| "learning_rate": 0.00012605246321177222, | |
| "loss": 1.1975, | |
| "step": 13600 | |
| }, | |
| { | |
| "epoch": 8.765195137555983, | |
| "grad_norm": 2.8301174640655518, | |
| "learning_rate": 0.00012477287268074214, | |
| "loss": 1.2049, | |
| "step": 13700 | |
| }, | |
| { | |
| "epoch": 8.829174664107486, | |
| "grad_norm": 2.866494655609131, | |
| "learning_rate": 0.00012350607805502239, | |
| "loss": 1.1916, | |
| "step": 13800 | |
| }, | |
| { | |
| "epoch": 8.89315419065899, | |
| "grad_norm": 3.0079123973846436, | |
| "learning_rate": 0.0001222264875239923, | |
| "loss": 1.1894, | |
| "step": 13900 | |
| }, | |
| { | |
| "epoch": 8.957133717210493, | |
| "grad_norm": 2.8487589359283447, | |
| "learning_rate": 0.00012094689699296224, | |
| "loss": 1.2077, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 9.021113243761997, | |
| "grad_norm": 2.8493270874023438, | |
| "learning_rate": 0.00011966730646193218, | |
| "loss": 1.1668, | |
| "step": 14100 | |
| }, | |
| { | |
| "epoch": 9.0850927703135, | |
| "grad_norm": 2.9474496841430664, | |
| "learning_rate": 0.0001183877159309021, | |
| "loss": 1.035, | |
| "step": 14200 | |
| }, | |
| { | |
| "epoch": 9.149072296865004, | |
| "grad_norm": 2.907160758972168, | |
| "learning_rate": 0.00011710812539987204, | |
| "loss": 1.0505, | |
| "step": 14300 | |
| }, | |
| { | |
| "epoch": 9.213051823416507, | |
| "grad_norm": 2.9124813079833984, | |
| "learning_rate": 0.00011582853486884196, | |
| "loss": 1.0573, | |
| "step": 14400 | |
| }, | |
| { | |
| "epoch": 9.277031349968011, | |
| "grad_norm": 2.925597906112671, | |
| "learning_rate": 0.00011454894433781189, | |
| "loss": 1.0654, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 9.341010876519514, | |
| "grad_norm": 2.844844102859497, | |
| "learning_rate": 0.00011326935380678182, | |
| "loss": 1.0703, | |
| "step": 14600 | |
| }, | |
| { | |
| "epoch": 9.404990403071018, | |
| "grad_norm": 3.0153064727783203, | |
| "learning_rate": 0.00011198976327575174, | |
| "loss": 1.0727, | |
| "step": 14700 | |
| }, | |
| { | |
| "epoch": 9.46896992962252, | |
| "grad_norm": 3.0144588947296143, | |
| "learning_rate": 0.00011071017274472169, | |
| "loss": 1.0791, | |
| "step": 14800 | |
| }, | |
| { | |
| "epoch": 9.532949456174025, | |
| "grad_norm": 3.219855308532715, | |
| "learning_rate": 0.0001094305822136916, | |
| "loss": 1.0921, | |
| "step": 14900 | |
| }, | |
| { | |
| "epoch": 9.596928982725528, | |
| "grad_norm": 3.0687131881713867, | |
| "learning_rate": 0.00010815099168266154, | |
| "loss": 1.0968, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 9.660908509277032, | |
| "grad_norm": 2.7260611057281494, | |
| "learning_rate": 0.00010687140115163146, | |
| "loss": 1.0924, | |
| "step": 15100 | |
| }, | |
| { | |
| "epoch": 9.724888035828535, | |
| "grad_norm": 3.266075372695923, | |
| "learning_rate": 0.0001055918106206014, | |
| "loss": 1.0871, | |
| "step": 15200 | |
| }, | |
| { | |
| "epoch": 9.78886756238004, | |
| "grad_norm": 2.816058874130249, | |
| "learning_rate": 0.00010431222008957132, | |
| "loss": 1.1183, | |
| "step": 15300 | |
| }, | |
| { | |
| "epoch": 9.852847088931542, | |
| "grad_norm": 2.7959651947021484, | |
| "learning_rate": 0.00010303262955854126, | |
| "loss": 1.1188, | |
| "step": 15400 | |
| }, | |
| { | |
| "epoch": 9.916826615483046, | |
| "grad_norm": 2.996344566345215, | |
| "learning_rate": 0.00010175303902751119, | |
| "loss": 1.1112, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 9.980806142034549, | |
| "grad_norm": 3.110300302505493, | |
| "learning_rate": 0.00010047344849648112, | |
| "loss": 1.13, | |
| "step": 15600 | |
| }, | |
| { | |
| "epoch": 10.044785668586053, | |
| "grad_norm": 2.8949873447418213, | |
| "learning_rate": 9.919385796545104e-05, | |
| "loss": 1.0038, | |
| "step": 15700 | |
| }, | |
| { | |
| "epoch": 10.108765195137556, | |
| "grad_norm": 2.7868170738220215, | |
| "learning_rate": 9.791426743442099e-05, | |
| "loss": 0.9668, | |
| "step": 15800 | |
| }, | |
| { | |
| "epoch": 10.17274472168906, | |
| "grad_norm": 3.0076348781585693, | |
| "learning_rate": 9.663467690339091e-05, | |
| "loss": 0.9592, | |
| "step": 15900 | |
| }, | |
| { | |
| "epoch": 10.236724248240563, | |
| "grad_norm": 3.313863515853882, | |
| "learning_rate": 9.535508637236083e-05, | |
| "loss": 0.9923, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 10.300703774792067, | |
| "grad_norm": 3.0162906646728516, | |
| "learning_rate": 9.407549584133077e-05, | |
| "loss": 0.9932, | |
| "step": 16100 | |
| }, | |
| { | |
| "epoch": 10.36468330134357, | |
| "grad_norm": 3.1555402278900146, | |
| "learning_rate": 9.279590531030069e-05, | |
| "loss": 0.9909, | |
| "step": 16200 | |
| }, | |
| { | |
| "epoch": 10.428662827895074, | |
| "grad_norm": 2.6180832386016846, | |
| "learning_rate": 9.151631477927062e-05, | |
| "loss": 1.0052, | |
| "step": 16300 | |
| }, | |
| { | |
| "epoch": 10.492642354446577, | |
| "grad_norm": 3.2279481887817383, | |
| "learning_rate": 9.023672424824054e-05, | |
| "loss": 1.0014, | |
| "step": 16400 | |
| }, | |
| { | |
| "epoch": 10.556621880998081, | |
| "grad_norm": 3.1148877143859863, | |
| "learning_rate": 8.895713371721049e-05, | |
| "loss": 1.0179, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 10.620601407549584, | |
| "grad_norm": 2.617116928100586, | |
| "learning_rate": 8.767754318618041e-05, | |
| "loss": 1.0234, | |
| "step": 16600 | |
| }, | |
| { | |
| "epoch": 10.684580934101088, | |
| "grad_norm": 3.179914951324463, | |
| "learning_rate": 8.639795265515034e-05, | |
| "loss": 1.018, | |
| "step": 16700 | |
| }, | |
| { | |
| "epoch": 10.748560460652591, | |
| "grad_norm": 3.002013921737671, | |
| "learning_rate": 8.511836212412028e-05, | |
| "loss": 1.03, | |
| "step": 16800 | |
| }, | |
| { | |
| "epoch": 10.812539987204095, | |
| "grad_norm": 3.1604723930358887, | |
| "learning_rate": 8.383877159309021e-05, | |
| "loss": 1.0289, | |
| "step": 16900 | |
| }, | |
| { | |
| "epoch": 10.876519513755598, | |
| "grad_norm": 3.0055463314056396, | |
| "learning_rate": 8.255918106206013e-05, | |
| "loss": 1.0244, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 10.940499040307103, | |
| "grad_norm": 3.2063984870910645, | |
| "learning_rate": 8.127959053103007e-05, | |
| "loss": 1.0246, | |
| "step": 17100 | |
| }, | |
| { | |
| "epoch": 11.004478566858605, | |
| "grad_norm": 2.9076807498931885, | |
| "learning_rate": 7.999999999999999e-05, | |
| "loss": 1.0284, | |
| "step": 17200 | |
| }, | |
| { | |
| "epoch": 11.06845809341011, | |
| "grad_norm": 2.873387098312378, | |
| "learning_rate": 7.872040946896993e-05, | |
| "loss": 0.8981, | |
| "step": 17300 | |
| }, | |
| { | |
| "epoch": 11.132437619961612, | |
| "grad_norm": 3.000307083129883, | |
| "learning_rate": 7.744081893793986e-05, | |
| "loss": 0.8994, | |
| "step": 17400 | |
| }, | |
| { | |
| "epoch": 11.196417146513117, | |
| "grad_norm": 2.965081214904785, | |
| "learning_rate": 7.616122840690978e-05, | |
| "loss": 0.9159, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 11.26039667306462, | |
| "grad_norm": 2.9889376163482666, | |
| "learning_rate": 7.488163787587971e-05, | |
| "loss": 0.9196, | |
| "step": 17600 | |
| }, | |
| { | |
| "epoch": 11.324376199616124, | |
| "grad_norm": 3.2027907371520996, | |
| "learning_rate": 7.360204734484964e-05, | |
| "loss": 0.9258, | |
| "step": 17700 | |
| }, | |
| { | |
| "epoch": 11.388355726167626, | |
| "grad_norm": 2.8633768558502197, | |
| "learning_rate": 7.233525271912987e-05, | |
| "loss": 0.9217, | |
| "step": 17800 | |
| }, | |
| { | |
| "epoch": 11.45233525271913, | |
| "grad_norm": 3.171734571456909, | |
| "learning_rate": 7.10556621880998e-05, | |
| "loss": 0.92, | |
| "step": 17900 | |
| }, | |
| { | |
| "epoch": 11.516314779270633, | |
| "grad_norm": 3.0337626934051514, | |
| "learning_rate": 6.977607165706973e-05, | |
| "loss": 0.9361, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 11.580294305822138, | |
| "grad_norm": 3.2068841457366943, | |
| "learning_rate": 6.849648112603966e-05, | |
| "loss": 0.9255, | |
| "step": 18100 | |
| }, | |
| { | |
| "epoch": 11.64427383237364, | |
| "grad_norm": 3.1199960708618164, | |
| "learning_rate": 6.721689059500959e-05, | |
| "loss": 0.9362, | |
| "step": 18200 | |
| }, | |
| { | |
| "epoch": 11.708253358925145, | |
| "grad_norm": 3.16876220703125, | |
| "learning_rate": 6.593730006397953e-05, | |
| "loss": 0.9509, | |
| "step": 18300 | |
| }, | |
| { | |
| "epoch": 11.772232885476647, | |
| "grad_norm": 2.9640047550201416, | |
| "learning_rate": 6.465770953294944e-05, | |
| "loss": 0.9338, | |
| "step": 18400 | |
| }, | |
| { | |
| "epoch": 11.836212412028152, | |
| "grad_norm": 3.112344980239868, | |
| "learning_rate": 6.337811900191938e-05, | |
| "loss": 0.9417, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 11.900191938579654, | |
| "grad_norm": 3.0994222164154053, | |
| "learning_rate": 6.211132437619961e-05, | |
| "loss": 0.9573, | |
| "step": 18600 | |
| }, | |
| { | |
| "epoch": 11.964171465131159, | |
| "grad_norm": 3.336512327194214, | |
| "learning_rate": 6.0831733845169535e-05, | |
| "loss": 0.9523, | |
| "step": 18700 | |
| }, | |
| { | |
| "epoch": 12.028150991682661, | |
| "grad_norm": 2.5776188373565674, | |
| "learning_rate": 5.955214331413947e-05, | |
| "loss": 0.9043, | |
| "step": 18800 | |
| }, | |
| { | |
| "epoch": 12.092130518234166, | |
| "grad_norm": 2.4695932865142822, | |
| "learning_rate": 5.8272552783109394e-05, | |
| "loss": 0.8316, | |
| "step": 18900 | |
| }, | |
| { | |
| "epoch": 12.156110044785668, | |
| "grad_norm": 2.912343740463257, | |
| "learning_rate": 5.699296225207933e-05, | |
| "loss": 0.8499, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 12.220089571337173, | |
| "grad_norm": 2.9618356227874756, | |
| "learning_rate": 5.571337172104926e-05, | |
| "loss": 0.8406, | |
| "step": 19100 | |
| }, | |
| { | |
| "epoch": 12.284069097888676, | |
| "grad_norm": 2.899482011795044, | |
| "learning_rate": 5.4433781190019186e-05, | |
| "loss": 0.8472, | |
| "step": 19200 | |
| }, | |
| { | |
| "epoch": 12.34804862444018, | |
| "grad_norm": 3.3630785942077637, | |
| "learning_rate": 5.315419065898912e-05, | |
| "loss": 0.8585, | |
| "step": 19300 | |
| }, | |
| { | |
| "epoch": 12.412028150991683, | |
| "grad_norm": 3.245290517807007, | |
| "learning_rate": 5.187460012795905e-05, | |
| "loss": 0.8593, | |
| "step": 19400 | |
| }, | |
| { | |
| "epoch": 12.476007677543187, | |
| "grad_norm": 3.1242306232452393, | |
| "learning_rate": 5.059500959692898e-05, | |
| "loss": 0.8659, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 12.53998720409469, | |
| "grad_norm": 3.265775442123413, | |
| "learning_rate": 4.931541906589891e-05, | |
| "loss": 0.879, | |
| "step": 19600 | |
| }, | |
| { | |
| "epoch": 12.603966730646194, | |
| "grad_norm": 2.995530843734741, | |
| "learning_rate": 4.8035828534868836e-05, | |
| "loss": 0.8688, | |
| "step": 19700 | |
| }, | |
| { | |
| "epoch": 12.667946257197697, | |
| "grad_norm": 3.1246113777160645, | |
| "learning_rate": 4.675623800383877e-05, | |
| "loss": 0.8759, | |
| "step": 19800 | |
| }, | |
| { | |
| "epoch": 12.731925783749201, | |
| "grad_norm": 3.249753713607788, | |
| "learning_rate": 4.54766474728087e-05, | |
| "loss": 0.8663, | |
| "step": 19900 | |
| }, | |
| { | |
| "epoch": 12.795905310300704, | |
| "grad_norm": 3.2970869541168213, | |
| "learning_rate": 4.419705694177863e-05, | |
| "loss": 0.8618, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 12.859884836852208, | |
| "grad_norm": 3.212738275527954, | |
| "learning_rate": 4.291746641074856e-05, | |
| "loss": 0.8709, | |
| "step": 20100 | |
| }, | |
| { | |
| "epoch": 12.92386436340371, | |
| "grad_norm": 3.064932107925415, | |
| "learning_rate": 4.163787587971848e-05, | |
| "loss": 0.8756, | |
| "step": 20200 | |
| }, | |
| { | |
| "epoch": 12.987843889955215, | |
| "grad_norm": 3.0357518196105957, | |
| "learning_rate": 4.035828534868841e-05, | |
| "loss": 0.8792, | |
| "step": 20300 | |
| }, | |
| { | |
| "epoch": 13.051823416506718, | |
| "grad_norm": 2.5418007373809814, | |
| "learning_rate": 3.9078694817658345e-05, | |
| "loss": 0.8018, | |
| "step": 20400 | |
| }, | |
| { | |
| "epoch": 13.115802943058222, | |
| "grad_norm": 3.088637590408325, | |
| "learning_rate": 3.779910428662827e-05, | |
| "loss": 0.7951, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 13.179782469609725, | |
| "grad_norm": 3.0088999271392822, | |
| "learning_rate": 3.6519513755598204e-05, | |
| "loss": 0.7759, | |
| "step": 20600 | |
| }, | |
| { | |
| "epoch": 13.24376199616123, | |
| "grad_norm": 2.929150104522705, | |
| "learning_rate": 3.523992322456814e-05, | |
| "loss": 0.7967, | |
| "step": 20700 | |
| }, | |
| { | |
| "epoch": 13.307741522712732, | |
| "grad_norm": 2.795482873916626, | |
| "learning_rate": 3.396033269353806e-05, | |
| "loss": 0.8023, | |
| "step": 20800 | |
| }, | |
| { | |
| "epoch": 13.371721049264236, | |
| "grad_norm": 2.998296022415161, | |
| "learning_rate": 3.2680742162507996e-05, | |
| "loss": 0.7929, | |
| "step": 20900 | |
| }, | |
| { | |
| "epoch": 13.435700575815739, | |
| "grad_norm": 3.09736967086792, | |
| "learning_rate": 3.140115163147792e-05, | |
| "loss": 0.801, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 13.499680102367243, | |
| "grad_norm": 3.1578280925750732, | |
| "learning_rate": 3.0134357005758153e-05, | |
| "loss": 0.8051, | |
| "step": 21100 | |
| }, | |
| { | |
| "epoch": 13.563659628918746, | |
| "grad_norm": 3.167719841003418, | |
| "learning_rate": 2.8854766474728086e-05, | |
| "loss": 0.8035, | |
| "step": 21200 | |
| }, | |
| { | |
| "epoch": 13.62763915547025, | |
| "grad_norm": 3.2616937160491943, | |
| "learning_rate": 2.7575175943698016e-05, | |
| "loss": 0.8152, | |
| "step": 21300 | |
| }, | |
| { | |
| "epoch": 13.691618682021753, | |
| "grad_norm": 3.2310030460357666, | |
| "learning_rate": 2.6295585412667945e-05, | |
| "loss": 0.8136, | |
| "step": 21400 | |
| }, | |
| { | |
| "epoch": 13.755598208573257, | |
| "grad_norm": 3.1108453273773193, | |
| "learning_rate": 2.5015994881637874e-05, | |
| "loss": 0.8158, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 13.81957773512476, | |
| "grad_norm": 2.875555992126465, | |
| "learning_rate": 2.37364043506078e-05, | |
| "loss": 0.8203, | |
| "step": 21600 | |
| }, | |
| { | |
| "epoch": 13.883557261676264, | |
| "grad_norm": 2.9165468215942383, | |
| "learning_rate": 2.2456813819577733e-05, | |
| "loss": 0.8142, | |
| "step": 21700 | |
| }, | |
| { | |
| "epoch": 13.947536788227767, | |
| "grad_norm": 2.839167356491089, | |
| "learning_rate": 2.1177223288547663e-05, | |
| "loss": 0.8075, | |
| "step": 21800 | |
| }, | |
| { | |
| "epoch": 14.011516314779271, | |
| "grad_norm": 2.6591968536376953, | |
| "learning_rate": 1.9897632757517592e-05, | |
| "loss": 0.7897, | |
| "step": 21900 | |
| }, | |
| { | |
| "epoch": 14.075495841330774, | |
| "grad_norm": 2.895771026611328, | |
| "learning_rate": 1.861804222648752e-05, | |
| "loss": 0.7396, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 14.139475367882278, | |
| "grad_norm": 3.234828472137451, | |
| "learning_rate": 1.733845169545745e-05, | |
| "loss": 0.7564, | |
| "step": 22100 | |
| }, | |
| { | |
| "epoch": 14.203454894433781, | |
| "grad_norm": 3.1565563678741455, | |
| "learning_rate": 1.605886116442738e-05, | |
| "loss": 0.7555, | |
| "step": 22200 | |
| }, | |
| { | |
| "epoch": 14.267434420985285, | |
| "grad_norm": 2.6395761966705322, | |
| "learning_rate": 1.4779270633397312e-05, | |
| "loss": 0.752, | |
| "step": 22300 | |
| }, | |
| { | |
| "epoch": 14.331413947536788, | |
| "grad_norm": 2.860470771789551, | |
| "learning_rate": 1.3499680102367243e-05, | |
| "loss": 0.7592, | |
| "step": 22400 | |
| }, | |
| { | |
| "epoch": 14.395393474088293, | |
| "grad_norm": 3.0977256298065186, | |
| "learning_rate": 1.222008957133717e-05, | |
| "loss": 0.7538, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 14.459373000639795, | |
| "grad_norm": 3.199491024017334, | |
| "learning_rate": 1.09404990403071e-05, | |
| "loss": 0.7554, | |
| "step": 22600 | |
| }, | |
| { | |
| "epoch": 14.5233525271913, | |
| "grad_norm": 3.1228437423706055, | |
| "learning_rate": 9.660908509277031e-06, | |
| "loss": 0.7536, | |
| "step": 22700 | |
| }, | |
| { | |
| "epoch": 14.587332053742802, | |
| "grad_norm": 3.0725913047790527, | |
| "learning_rate": 8.38131797824696e-06, | |
| "loss": 0.745, | |
| "step": 22800 | |
| }, | |
| { | |
| "epoch": 14.651311580294307, | |
| "grad_norm": 2.7372710704803467, | |
| "learning_rate": 7.101727447216891e-06, | |
| "loss": 0.7761, | |
| "step": 22900 | |
| }, | |
| { | |
| "epoch": 14.71529110684581, | |
| "grad_norm": 2.8115973472595215, | |
| "learning_rate": 5.822136916186819e-06, | |
| "loss": 0.7549, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 14.779270633397314, | |
| "grad_norm": 3.0110971927642822, | |
| "learning_rate": 4.5425463851567495e-06, | |
| "loss": 0.7585, | |
| "step": 23100 | |
| }, | |
| { | |
| "epoch": 14.843250159948816, | |
| "grad_norm": 3.1187989711761475, | |
| "learning_rate": 3.2629558541266794e-06, | |
| "loss": 0.7612, | |
| "step": 23200 | |
| }, | |
| { | |
| "epoch": 14.90722968650032, | |
| "grad_norm": 3.022102117538452, | |
| "learning_rate": 1.983365323096609e-06, | |
| "loss": 0.7555, | |
| "step": 23300 | |
| }, | |
| { | |
| "epoch": 14.971209213051823, | |
| "grad_norm": 3.1614649295806885, | |
| "learning_rate": 7.037747920665386e-07, | |
| "loss": 0.7546, | |
| "step": 23400 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 23445, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 15, | |
| "save_steps": 2500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 9.649122201906708e+17, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |